The Raw Milk Microbiota from Semi-Subsistence Farms Characteristics by NGS Analysis Method

molecules

Article The Raw Milk Microbiota from Semi-Subsistence Farms Characteristics by NGS Analysis Method

Bartosz Hornik 1 , Jakub Czarny 1 , Justyna Staninska-Pi˛eta 2 , Łukasz Wolko 3 , Paweł Cyplik 4 and Agnieszka Piotrowska-Cyplik 2,*

1 Institute of Forensic Genetics, Al. Mickiewicza 3/4, 85-071 Bydgoszcz, Poland; [email protected] (B.H.); [email protected] (J.C.) 2 Department of Food Technology of Plant Origin, Poznan University of Life Sciences, Wojska Polskiego 31, 60-624 Poznań,Poland; [email protected] 3 Department of Biochemistry and Biotechnology, Poznan University of Life Sciences, Dojazd 11, 60-632 Poznań,Poland; [email protected] 4 Department Biotechnology and Food Microbiology, Poznan University of Life Sciences, Wojska Polskiego 48, 60-627 Poznań,Poland; [email protected] * Correspondence: [email protected]; Tel.: +48-618487284

Abstract: The aim of this study was to analyze the microbiome of raw milk obtained from three semi- subsistence farms (A, B, and C) located in the Kuyavian-Pomeranian Voivodeship in Poland. The composition of drinking milk was assessed on the basis of 16S rRNA gene sequencing using the Ion Torrent platform. Based on the conducted research, significant changes in the composition of the milk microbiome were found depending on its place of origin. Bacteria belonging to the Bacillus (17.0%), Citation: Hornik, B.; Czarny, J.; Corynebacterium (12.0%) and Escherichia-Shigella (11.0%) genera were dominant in the milk collected Staninska-Pi˛eta,J.; Wolko, Ł.; from farm A. In the case of the milk from farm B, the dominant bacteria belonged to the Acinetobacter Cyplik, P.; Piotrowska-Cyplik, A. The genus (21.0%), whereas in the sample from farm C, Escherichia-Shigella (24.8%) and Bacillus (10.3%) Raw Milk Microbiota from Semi- dominated the microbiome. An analysis was performed using the PICRUSt tool (Phylogenetic Subsistence Farms Characteristics by Investigation of Communities by Reconstruction of Unobserved States) in order to generate a profile NGS Analysis Method. Molecules 2021, 26, 5029. https://doi.org/ of genes responsible for bacterial metabolism. The conducted analysis confirmed the diversity of the 10.3390/molecules26165029 profile of genes responsible for bacterial metabolism in all the tested samples. On the other hand, simultaneous analysis of six KEGG Orthologs (KO), which participated in beta-lactam resistance Academic Editors: Antonio- responsible for antibiotic resistance of bacteria, demonstrated that there is no significant relationship José Trujillo, Radmila Pavlovic, between the predicted occurrence of these orthologs and the place of existence of microorganisms. Luca Chiesa, Sara Panseri and Therefore, it can be supposed that bacterial resistance to beta-lactam antibiotics occurs regardless of Emanuela Zanardi the environmental niche, and that the antibiotic resistance maintained in the population is a factor that shapes the functional structure of the microbial consortia. Received: 23 June 2021 Accepted: 16 August 2021 Keywords: antibiotic resistance; raw milk microbiome; farm; next-generation sequencing Published: 19 August 2021

Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in 1. Introduction published maps and institutional affil- iations. Milk and its products play an important role in human nutrition in many cultures. It is a source of protein, vitamins and many minerals, which makes it an excellent environment for the growth and development of bacteria [1,2]. Raw cow milk is one of the most diverse raw materials in terms of microbiology, which directly affects the quality and price of manufactured products and the company’s financial results [3]. The colonization of milk Copyright: © 2021 by the authors. Licensee MDPI, Basel, Switzerland. by microorganisms is a significant threat that has a negative impact not only on the quality This article is an open access article and durability of products, but also on human health [4]. In order to achieve a positive distributed under the terms and effect of the health properties of milk on the human body, appropriate hygienic conditions conditions of the Creative Commons should be maintained not only during collection, but also during its transport to dairy Attribution (CC BY) license (https:// plants, processing and preservation. creativecommons.org/licenses/by/ Milk, in addition to its endogenous microbiota, enables the development of microor- 4.0/). ganisms which may originate from the surface of animals, as well as from the environment

Molecules 2021, 26, 5029. https://doi.org/10.3390/molecules26165029 https://www.mdpi.com/journal/molecules Molecules 2021, 26, 5029 2 of 12

in which the livestock lives [4]. The contamination of milk with bacteria mainly results from poorly cleaned and poorly disinfected milking equipment, lack of hygiene during milking or handling of the milk, and bacteria present in the barn. Moreover, the presence of environmental pathogens such as Escherichia coli, Klebsiella spp., Streptococcus dysgalactiae and Streptococcus uberis may contribute to the development of mastitis in dairy cattle, which is an “occupational disease” of high-yielding dairy cows [5,6]. The disease is caused by bacteria and other microorganisms that enter the teats. The appearance of this disease in high-yielding dairy cows is the cause of huge economic losses [7]. There are two forms of mastitis: clinical and subclinical, which is a latent form. It is estimated that up to 50% of cows may suffer from subclinical mastitis. Unfortunately, in the subclinical form, there are also economic losses caused by the reduced milk yield of cows; hence, the correct diagnosis and implementation of an effective treatment process are very important [8]. The treatment of mastitis in cows, despite the implementation of prophylactic pro- grams and new methods of therapy, is still mainly based on the administration of antibiotics. It is estimated that animal production in most countries sometimes accounts for as much as 80% of the total consumption of antibiotics [9]. Unfortunately, their presence in food has negative health and economic consequences for humans [10]. One of the greatest threats to public and global health associated with the use of antibiotics is the increase in antibiotic resistance of bacteria [11]. It is assumed that the main cause of this phenomenon is their excessive use, and the scale of abuse in this area makes them the main cause of the growth and spread of antibiotic-resistant bacteria and resistance genes in the environment [12]. Antibiotics used to treat and prevent bacterial infections in animals may also contribute to the formation of drug-resistant bacterial strains in the human body [13]. It can be said with certainty that the process of managing antibiotic resistance has progressed considerably, meaning that it is leading the world into the post-antibiotic era. Unfortunately, the alarming forecasts from scientists have become completely realistic, and harmless bacterial infections that have been successfully treated in the past appear to be deadly today [14]. Among the antibiotics, more than 60% of the intramammary preparations used contain β-lactams (penicillins, cephalosporins), which is confirmed by the results of the presence of antibiotics in milk. One of the main reasons for the insensitivity to antibiotics is the ability of bacteria to produce β-lactamase, an enzyme that neutralizes the action of, for example, penicillins and cephalosporins [12]. On the one hand, the subthreshold concentrations of antibiotics strongly influence the selection of resistant strains, and they contribute to the formation of morphologically changed bacteria [15,16]. In addition, the environment in which antibiotics are present favors the transfer of mobile genetic elements by horizontal gene transfer, which leads to the dissemination of resistance genes even between phylogenetically distant bacteria [17]. Due to the above-mentioned reasons, the assessment of the composition and direction of development of the microbiota present in raw milk has a significant impact not only on the composition and quality of milk, but above all on the quality of dairy products [18]. The microorganisms present in these products may affect human health. They constitute a reservoir of genes that determine the antibiotic resistance of bacteria, which can be permanently transferred to the microbiome of the human gastrointestinal tract [19]. On this basis, sequence analysis of the hypervariable regions of the 16S rRNA gene using the Ion Torrent platform (Life Technologies, Carlsbad, CA, USA) was used to assess the microbiome of milk from semi-subsistence farms. Then, a functional analysis of the milk microbiome was carried out based on the PICRUSt tool in order to determine the potential of the microbiome as a carrier of genes that determine the resistance of bacteria to β-lactam antibiotics. Molecules 2021, 26, x 3 of 13

Molecules 2021, 26, 5029 3 of 12

2. Results and Discussion 2.1. Taxonomic Analysis of the Milk Microbiome Milk is is an an excellent excellent environment environment for for the the development development of of microorganisms microorganisms responsi- respon- blesible for for the the specific specific properties properties of ofmany many dairy dairy products products as well as well as microbes as microbes which which are un- are desirableundesirable for fortechnological technological and andhealth health reasons reasons (review (review by Quigley by Quigley et al. et[3]). al. The [3]). use The of next-generationuse of next-generation sequencing sequencing allowed allowed to detect to DNA detect from DNA bacteria, from bacteria,the presence the presenceof which couldof which not couldbe confirmed not be confirmed to date by tocultivat date byion-based cultivation-based methods [17,20,21]. methods [The17,20 taxonomic,21]. The identificationtaxonomic identification carried out on carried the basis out onof th thee sequence basis of theanalysis sequence of the analysis hypervariable of the hy-re- gionspervariable of the regions16S rRNA of gene the 16S based rRNA on the gene SI basedLVA v119 on thedatabase SILVA enabled v119 database the detection enabled of microorganismsthe detection of microorganismsthat are components that of are milk components collected offrom milk cows collected from semi-subsistence from cows from farms.semi-subsistence farms. In all of the studied samples, Firmicutes was the dominant phylum and its ratio in the milk microbiome ranged from 42% to 53%. The other types present in milkmilk includedincluded Proteobacteria (40–26%), Actionobacteria (14–5%) andand BacteroidetesBacteroidetes (10–3%).(10–3%). All samplessamples displayed the presence of bacteria belonging to 42 classes (Figure1 1).).

Figure 1. The relativerelative abundancesabundances of bacterial classes in milk collected from semi-subsistence farms (A1—A10, B1—B10, C1—C10 representrepresent the the individual individual samples samples from from respective respective farms; farms; A, B, CA, represent B, C represent the mean the value mean for allvalue samples for all for samples each farm). for each farm). Five classes of bacteria were found to be dominant in the milk samples collected from all ofFive the researchedclasses of bacteria farms: Gammaproteobacteria were found to be dominant (21.93–42.77%), in the milk Bacilli samples (17.88–36.26%), collected fromClostridia all (7.04–14.88%),of the researched Actinobacteria farms: (5.14–13.82%)Gammaproteobacteria and Betaproteobacteria (21.93–42.77%), (1.56–7.28%). Bacilli (17.88–36.26%), Clostridia (7.04–14.88%), Actinobacteria (5.14–13.82%) and Betaproteo- bacteria2.2. Lactic (1.56–7.28%). Fermentation Bacteria The newly carried out generation sequencing allowed for species and functional identi- 2.2.ﬁcation Lactic of Fermentation lactic acid bacteria Bacteria (LAB). In all samples, different amounts of lactic acid bacteria (LAB)The were newly found, carried which out belonged generation to the sequencing following allowed types: Carnobacterium for species and, Enterococcus functional, identificationLacticigenium, ofLactobacillus lactic acid ,bacteriaLactococcus (LAB)., Leuconostoc In all samples,, Streptococcus different andamountsTrichococcus of lactic. acid The bacteriahighest ratio(LAB) of were LAB found, bacteria which in the belonged milk microbiome to the following was found types: in the Carnobacterium sample collected, En- from farm C and it was equal to 9.62%. In other farms, it ranged from 5.2% to 6.1%.

Molecules 2021, 26, 5029 4 of 12

LABs are the most important group of microorganisms found in milk, because their role is to convert carbohydrates and proteins into numerous secondary metabolites [9]. The compounds produced by these bacteria stimulate the growth of other groups of microorganisms, and thus directly and indirectly determine the ﬁnal shape of the ﬁnished product [22].

2.3. Spoiled Milk Bacteria Some scientists have suggested that the bacteria found in milk originate not only from external sources, but also enter the milk as a result of migration from other internal organs through the so-called internal colonization. The presence of bacteria belonging to the Ruminococcus and Bifidobacterium genera, as well as to the Peptostreptococcaceae family, in the analyzed milk samples was first described by Young et al. [23], who showed in their research that these bacteria can escape from the intestinal lumen and travel through the mesentery lymph nodes to the mammary gland. The higher sensitivity of the used molecular methods compared to the cultivation- based methods revealed the presence of numerous microorganisms as responsible for the spoilage of milk. A large group identified in all of the analyzed milk samples were Gram- negative bacteria belonging to the Pseudomonas, Acinetobacter and Aeromonas genera [24]. Such bacteria are characterized by the ability to produce lipases, which are responsible for unfavorable changes in the taste and smell of milk. The milk also contains coliform bacteria, which are characterized by the ability to ferment lactose, and Gram-positive spore bacteria (Bacillus) responsible for milk coagulation. The presence of these groups of bacteria is confirmed by numerous previous studies of the milk microbiota [20,21]. Studies of the milk microbiota based on next-generation sequencing also identified anaerobic bacteria such as Bacteroides, Faecalibacterium, Prevotella and Catenibacterium, the presence of which may be related to the presence of fecal contaminants.

2.4. Pathogenic Bacteria Causing Mastitis A large group of microorganisms present in milk samples included pathogenic and potentially pathogenic bacteria responsible for the occurrence of mastitis in dairy cattle. The identified pathogenic bacteria belonged to the Proteobacteria and Firmicutes phyla. There are several pathogens that most often cause inflammation of the mammary glands in cows: Staphylococcus aureus, Escherichia coli, Streptococcus agalactiae, Streptococcus dysgalactiae, Streptococcus uberis, coagulase-negative Staphylococci, Enterococcus spp. (mainly E. faecium and E. faecalis), and their prevalence differs in every country and even in every herd that are tested. Proteobacteria are a diverse taxonomic unit that covers a wide group of pathogens. They are Gram-negative bacteria considered to be environmental mastitis pathogens, in contrast to Gram-positive bacteria belonging to Firmicutes, which are considered contagious mastitis pathogens [25–27]. However, in terms of the generic composition of bacteria, significant differences were found between the analyzed samples collected from the researched farms (Figure2). Bacteria belonging to the Bacillus (17%), Corynebacterium (12%) and Escherichia-Shigella (11%) genera dominated in the milk collected from farm A. In the milk from farm B, bacteria belonging to the Acinetobacter genus were dominant (21%), whereas in the sample from farm C, Escherichia-Shigella (24.8%) and Bacillus (10.3%) were the most abundant members. The ratio of other types of bacteria did not exceed 10% in all of the samples collected from farms A, B and C. The dominant bacterium was Escherichia coli. Among the remaining 20 dominant bacteria, as many as 12 were uncultured bacteria, which cannot be identified and diagnosed using classical microbiology methods. The other dominant bacteria included: Achromobacter xylosoxidans, Streptococcus pyogenes, Streptococcus pseudoporcinus, Bacillus subtilis i Chryseobac- terium hagamense, Corynebacterium bovis, Aerococcus suis, and Corynebacterium sphenisci. Molecules 2021, 26, x 5 of 13

The dominant bacterium was Escherichia coli. Among the remaining 20 dominant bacteria, as many as 12 were uncultured bacteria, which cannot be identified and diagnosed using classical microbiology methods. The other dominant bacteria included: Achromobacter xylosoxidans, Streptococcus pyogenes, Streptococcus pseudoporcinus, Bacillus Molecules 2021, 26, x 5 of 13 Molecules 2021, 26, 5029 subtilis i Chryseobacterium hagamense, Corynebacterium bovis, Aerococcus suis5 of, 12 and Coryne- bacterium sphenisci. The dominant bacterium was Escherichia coli. Among the remaining 20 dominant bacteria, as many as 12 were uncultured bacteria, which cannot be identified and diagnosed using classical microbiology methods. The other dominant bacteria included: Achromobacter xylosoxidans, Streptococcus pyogenes, Streptococcus pseudoporcinus, Bacillus subtilis i Chryseobacterium hagamense, Corynebacterium bovis, Aerococcus suis, and Coryne- bacterium sphenisci.

Figure 2. The relative abundanceFigure 2.ofThe bacterial relativeFigure genera abundance 2. inThe milk of relative bacterial coll abundanceected genera from in of milk semi-subsistence bacterial collected genera from in semi-subsistence farmsmilk coll (A,ected B, C—milkfrom farms semi-subsistence from farms (A, B, C—milk from respective farms). —mean, —outlers point, ∗—extreme point respective farms). —mean,(A, —outlers B, C—milk point, from respective ∗—extreme farms). point—mean, —outlers point, ∗—extreme point.

Molecules 2021, 26, x 6 of 13

Molecules 2021, 26, 5029 2.5. Analysis of Biodiversity 6 of 12 The values of the analyzed alpha-biodiversity coefficients in the metapopulation of the microorganisms present in milk are presented in Table 1. Estimates of intrasample diversity2.5. Analysis were of Biodiversity carried out at a rarefaction depth of 100,000 reads per sample. There was a significantThe values difference of the analyzed in the alpha-biodiversitynumber of identified coefficients OTUs in the in metapopulationall tested samples. of Moreover, almostthe microorganisms all of the alpha-biodiversity present in milk are presented indices indiffer Table 1from. Estimates each ofother intrasample (except the Shannon diversity were carried out at a rarefaction depth of 100,000 reads per sample. There was a entropy,significant which difference showed in the numberno significant of identified differen OTUsce in between all tested samples.milk from Moreover, farm B and C). In all ofalmost the determined all of the alpha-biodiversity indicators, the indices microbiome differ from of eachmilk other obtained (except from the Shannon farm A displayed the highestentropy, whichbiodiversity. showed no significant difference between milk from farm B and C). In all of the determined indicators, the microbiome of milk obtained from farm A displayed the Tablehighest 1. biodiversity. Analysis of the alpha-biodiversity of bacteria present in the analyzed milk samples.

Table 1. Analysis of the alpha-biodiversityMilk from of bacteria Farm presentA Milk in the analyzedfrom Farm milk samples. B Milk from Farm C Number of OTUs Milk from 5490 Farm A± 78 Milk from Farm 4486 B ± 94 Milk from Farm C 3488 ± 58 Chao-1Number bias-corrected of OTUs 5490 ± 550378 ± 14 4486 ± 94 4597 ± 18 3488 ± 58 3508 ± 14 Chao-1Shannon bias-corrected entropy 5503 ± 8.5914 ± 0.45 4597 ± 18 7.02 ± 0.36 3508 ± 14 6.33 ± 0.39 Shannon entropy 8.59 ± 0.45 7.02 ± 0.36 6.33 ± 0.39 PhylogeneticPhylogenetic diversity 10.23 ± 0.61 8.98 ± 0.49 7.12 ± 0.44 10.23 ± 0.61 8.98 ± 0.49 7.12 ± 0.44 diversity The results of the beta-biodiversity analysis of variants collected from three farms are presentedThe results ofin theFigure beta-biodiversity 3. analysis of variants collected from three farms are presented in Figure3.

Figure 3. Principal coordinate analysis (PCoA) based on the Bray–Curtis dissimilarity met- Figurerics showing 3. Principal the distance coordinate in the bacterial analysis communities (PCoA) betweenbased on analyzed the Bray–Curtis samples (milk dissimilarity from metrics showingfarm A—A1, the A2 distance . . . ; milk fromin the farm ba B—B1,cterial B2 communities . . . ; milk from farmbetween C—C1, anal C2 .yzed . . .). samples (milk from farm A—A1, A2…; milk from farm B—B1, B2…; milk from farm C—C1, C2….). The same relationships were observed between the majority of the studied variables for both of the methods of scaling of three principal components in space. The location of pointsThe from same the individual relationships milk samples were observed in separate between spaces on thethe chart majority is consistent of the with studied variables fortheir both origin, of whichthe methods indicates of significant scaling differencesof three principal in the bacterial components composition in related space. to The location of pointsthe place from of their the collection. individual This milk is confirmed samples by thein predictionsseparate spaces of the functional on the chart metabolic is consistent with theirmetapopulation origin, which profiles indicates carried out significant overall (Figure differences4). in the bacterial composition related to the place of their collection. This is confirmed by the predictions of the functional metabolic metapopulation profiles carried out overall (Figure 4).

Molecules 2021, 26, x 7 of 13 Molecules 2021, 26, 5029 7 of 12

Figure 4. Clustering analysis of the overall, functional metabolic profile of bacterial communities in the analyzed milk Figure 4. Clustering analysis of the overall, functional metabolic profile of bacterial communities in the analyzed milk samples derived from farms A, B and C (a) principal coordinate analysis (PCA) and (b) dendrogram analysis showing the samples derived from farms A, B and C (a) principal coordinate analysis (PCA) and (b) dendrogram analysis showing the distance in the predicted functional metabolic profile between samples. distance in the predicted functional metabolic profile between samples. 2.6. Analysis of Functional Potential and Antibiotic Resistance 2.6. Analysis of Functional Potential and Antibiotic Resistance Based on the PCA analysis and the structure of the dendrogram, there are significant Based on the PCA analysis and the structure of the dendrogram, there are significant differences in the overall predicted functional potential between the populations of mi- differences in the overall predicted functional potential between the populations of microorganisms collected from the different farms. There is a clear influence of the ecological croorganisms collected from the different farms. There is a clear influence of the ecolog- niche on the shaping of the structure of the predicted (putative) bacterial metagenome [28]. ical nicheThe treatmenton the shaping of mastitis of the in cows,structure despite of the the predicted implementation (putative) of prophylactic bacterial meta- pro- genomegrams and[28]. new genetic methods of diagnosis and treatment (such as sequencing for identificationThe treatment of pathogens of mastitis and in screeningcows, despit fore antibiotic the implementation resistance genes),of prophylactic is still mainly pro- gramsbased and only new on the genetic administration methods of antibiotics,diagnosis and despite treatment many (such negative as sequencing aspects of their for identificationuse and inconsistent of pathogens satisfactory and screening effectiveness for antibiotic [29,30]. Oneresistance of the genes), main reasons is still mainly for the basedlack of only sensitivity on the administration to antibiotics is of the antibiotics, ability for despite bacteria many to produce negative enzymes aspects of (e.g., theirβ - uselactamase) and inconsistent that neutralize satisfactory the activity effectivene of selectedss [29,30]. antibiotics One of [31 the]. Themain genes reasons responsible for the lackfor theof productionsensitivity ofto βantibiotics-lactamases is are the found ability either for onbacteria the bacterial to produce chromosome enzymes or on(e.g., the βplasmid-lactamase) [32]. that neutralize the activity of selected antibiotics [31]. The genes responsible for theThe production production of ofβ-lactamases various β-lactamases are found either and acylases on the bybacterial Gram-positive chromosome and Gram-or on thenegative plasmid bacteria, [32]. which remove the side-substituent of the antibiotic, is the key cause of resistanceThe production to β-lactam of antibiotics various [33β-lactamases]. and acylases by Gram-positive and Gram-negativeThe variety bacteria, of bacterial which taxa, remove in which the the side presence-substituent of genes of the participating antibiotic, inis thethe beta-key causelactam of resistanceresistance pathwayto β-lactam was antibiotics predicted [33]. in this study, indicates the importance of the phenomenonThe variety of horizontalof bacterial transfer taxa, in of which this trait. the presence It also seems of genes that theparticipating features of in resistance the beta-lactamto beta-lactam resistance antibiotics pathway occur was regardless predicted of in the this environmental study, indicates niche the and importance are an important of the phenomenonfactor which shapesof horizontal the functional transfer structure of this oftrait. microbial It also consortia.seems that the features of re- sistanceThe to analysis, beta-lactam limited antibiotics to six KEGG occur Orthologsregardless (KO) of the that environmental participated inniche beta-lactam and are an re- importantsistance, showed factor which that there shapes is no the significant functional relationship structure of between microbial the consortia. predicted occurrence of theseThe orthologsanalysis, andlimited the placeto six ofKEGG microbial Orthologs existence (KO) (Figure that 5participated). A detailed, in predicted beta-lactam rep- resistance,resentation showed of these that genes there in individual is no signific phylaant and relationship bacterial classesbetween is presentedthe predicted in Figure occur-6 . renceThe functional of these orthologs features relatedand theto place resistance of microbial to beta-lactam existence antibiotics (Figure 5). are A detailed, not limited pre- to dicteda few representation taxonomic groups. of these Nevertheless, genes in indivi thedual following phyla phylaand bacterial in which classes the probability is presented of inthese Figure orthologs 6. The isfunctional particularly features high canrelated be distinguished: to resistance to Proteobacteria, beta-lactam antibiotics Firmicutes, are Bacte- not limitedrioidetes, to Actinobacteria,a few taxonomic Acidobacteria groups. Nevertheless, and Cyanobacteria. the following Moreover, phyla the in PCA which analysis the

Molecules 2021, 26, x 8 of 13

probability of these orthologs is particularly high can be distinguished: Proteobacteria, Firmicutes, Bacterioidetes, Actinobacteria, Acidobacteria and Cyanobacteria. Moreover, Molecules 2021, 26, 5029 probability of these orthologs is particularly high can be distinguished: Proteobacteria,8 of 12 the PCA analysis showed that beta-lactam resistance functional metabolic profiles did Firmicutes, Bacterioidetes, Actinobacteria, Acidobacteria and Cyanobacteria. Moreover, not differ significantly between the populations of microorganisms isolated from the the PCA analysis showed that beta-lactam resistance functional metabolic profiles did different farms (Figure 6). not differ significantly between the populations of microorganisms isolated from the showeddifferent that farms beta-lactam (Figure resistance6). functional metabolic proﬁles did not differ signiﬁcantly between the populations of microorganisms isolated from the different farms (Figure6).

Figure 5. The relative representation of beta-lactam resistance gene orthologs in (a) individual classes and (b) phyla of Figurebacteria 5. The in relative analyzed representation samples derived of beta-lactam from farmsresistance A, B and C. gene orthologs in (a) individual classes and (b) phyla of bacteriaFigure in5. analyzedThe relative samples representation derived from of beta-lactam farms A, B andresistance C. gene orthologs in (a) individual classes and (b) phyla of bacteria in analyzed samples derived from farms A, B and C.

Figure 6. Clustering analysis of beta-lactam resistance functional metabolic profile of bacterial communities in the analyzed milkFigure samples 6. derivedClustering from analysis farms A, of B beta and-lactam C: (a) principal resistance coordinate functional analysis metabolic (PCA) pr andofile (bof) dendrogrambacterial communities analysis showing in the ana- the distancelyzed milk in thesamples predicted derived functional from farms metabolic A, B and profile C: ( betweena) principal samples. coordinate analysis (PCA) and (b) dendrogram analysis Figureshowing 6. Clustering the distance analysis in the of predicted beta-lactam func resistancetional metabolic functional profile metabolic between pr samples.ofile of bacterial communities in the analyzed milk samples derived3. from Materials farms andA, B Methodsand C: (a) principal coordinate analysis (PCA) and (b) dendrogram analysis showing the distance in the predicted functional metabolic profile between samples. 3.1.3. Characteristics Materials and of FarmsMethods and Collection of Samples 3. 3.1.MaterialsThe Characteristics microflora and Methods of milkof Farms from and three Collection farms located of Samples in Poland in the Kuyavian-Pomeranian voivodship (farms A, B and C) was subjected to metagenomic analysis. The farms were 3.1. CharacteristicsThe microflora of Farms of andmilk Collection from threeof Samples farms located in Poland in the Kuyavi- semi-subsistencean-Pomeranian farms voivodship with 5 to(farms 8 dairy A, cows.B and AC) 500was mL subjected sample to of metagenomic the bulk milk analysis. was The microflora of milk from three farms located in Poland in the Kuyavi- collectedThe farms from were each semi-subsistence farm every week. farms The with samples 5 to collected8 dairy cows. during A 500 each mL month sample were of the an-Pomeranian voivodship (farms A, B and C) was subjected to metagenomic analysis. mixedbulk with milk each was other, collected and thus,from fromeach Februaryfarm every to Novemberweek. The samples 2020, 10 averagedcollected samplesduring each The farms were semi-subsistence farms with 5 to 8 dairy cows. A 500 mL sample of the frommonth each farmwere weremixed prepared. with each In other, total, and 30 samplesthus, from were February analyzed. to November The samples 2020, were 10 av- collectedbulk milk in was sterile collected containers from and each transported farm every to week. the microbiological The samples collected laboratory during at 4 ◦eachC, andmonth then were stored mixed at −20 with◦C. each other, and thus, from February to November 2020, 10 av-

Molecules 2021, 26, 5029 9 of 12

3.2. Isolation of DNA DNA isolation from milk samples was performed using the Genomic Mini AX Bacteria Spin kit (060-100S, A&A Biotechnology, Gdańsk,Poland) according to the protocol provided by the manufacturer. Finally, the purified DNA was eluted. The isolates were stored at −80 ◦C after they had been neutralized, in order to minimize matrix degradation. The efficiency of isolation was checked each time based on the fluorimetric method with the use of the Qbit 3.0 device and the Qubit™ dsDNA HS Assay Kit (Q32851, Ther- moFisher Scientific, Waltham, MA, USA). For each sample, three DNA extractions were performed and finally combined after a positive quantification.

3.3. PCR Amplification and NGS Sequencing The PCR reaction was prepared using the Ion 16S™ Metagenomics Kit (A26216, Life Technologies). This kit allows for the amplification of the V2–V9 regions of the bacterial 16S rRNA gene. The reaction was prepared according to the manufacturer’s instructions. The reaction consists of 15 µL of 2× Environmental Master Mix, 3 µL of the appropriate primer and 12 µL of the DNA sample previously isolated from the milk sample. The reaction was performed in a Veriti thermal cycler (Life Technologies) using the following temperature program: initial denaturation for 10 min at 95 ◦C; 25 cycles of denaturation for 30 s at 95 ◦C; annealing for 30 s at 58 ◦C; extension for 20 s at 72 ◦C; and a final extension for 7 min at 72 ◦C. The reaction products were purified using the Agencourt AMPure XP Reagent (A63880, Beckman Coulter, Pasadena, CA, USA), according to the manufacturer’s instructions. The method was based on binding DNA to magnetic beads followed by washing away the contaminants with ethanol. The DNA was rinsed from the beads using nuclease-free water or low-TE buffer. A library was prepared according to the manufacturer’s instructions using the Ion Plus Fragment Library Kit (4471252, Life Technologies). The prepared library was purified using Agencourt AMPure XP Reagent (A63880, Beckman Coulter, Pasadena, CA, USA), according to the manufacturer’s instructions. The concentration of the library was assessed using the Ion Universal Library Quantitation Kit and a real time PCR instrument—Quant Studio 5 (A26217, Life Technologies). The library was then diluted to a concentration of 10 pM. The diluted library was coated onto beads (used for sequencing) in emulsion PCR using the Ion PGM™ Hi-Q™ View OT2 Kit reagent kit and an Ion One Touch 2 Instrument (A29900, Life Technologies). The library-coated beads were purified using an Ion One Touch ES Instrument (Life Technologies). The library-coated beads were sequenced using an Ion PGM System (Life Technologies) using the Ion PGM™ Hi-Q™ View Sequencing Kit (A29900) on an Ion 316™ Chip Kit v2 BC. The 16S rRNA sequencing datasets generated and analyzed during the current study have been deposited at the National Center for Biotechnology Information (SRA repository), as BioProject under ID PRJNA699887.

3.4. Bioinformatic Analysis The sequence reads from the Ion Torrent (Thermo Fisher Scientiﬁc) in BAM format were imported into the CLC Genomics Workbench 20.0 software (Qiagen, Hilden, Germany) and processed with CLC Microbial Genomics Module 20.1.1 (Qiagen, Hilden, Germany). The total number of reads and results of downstream processing for all samples were presented in the Supplementary File (Table S1). Chimeric and low-quality reads (quality limit = 0.05, ambiguous limit = ‘N’) were ﬁltered and removed. Then, the sequence reads were clustered against the SILVA v119 [34] database at 97% similarity of operational taxonomic units (OTU). Finally, the merged abundance table was generated, and selected alpha (number of OTUs, Chao-1 bias-corrected, Shannon entropy and Phylogenetic diversity) and beta (Bray–Curtis principal coordinate analysis) diversity parameters were determined. Molecules 2021, 26, 5029 10 of 12

3.5. Prediction the Functional Profile from Targeted Metagenomic Data In order to generate a profile of the putative functional properties of the analyzed microbial consortia based on the targeted 16S rRNA OTU data, an analysis was carried out using the PICRUSt (ver.1) (Phylogenetic Investigation of Communities by Reconstruction of Unobserved States) tool [35]. The OTUs derived from clustering against GreenGenes 13.5 (97% similarity), which were normalized for 16S rRNA copy numbers, were used as input data. A total of 6909 KEGG Orthologs (KO) were identified (translate as abundance), 6 of which were related to beta lactam-resistance (Table2). PCA visualization and dendrogram analysis (Jensen–Shannon divergence) of the obtained data were performed using the MicrobiomeAnalyst software (McGill University, Montreal, Canada) [36].

Table 2. List of KEGG gene orthologs (KO) analyzed with the PICRUSt tool related to beta- lactam resistance.

KO Identiﬁer Gene Name K01467 beta-lactamase class C (ampC) K02171 BlaI family transcriptional regulator, penicillinase repressor (blaI) K02172 bla regulator protein blaR1 (blaR1) K02545 penicillin-binding protein 2 prime (mecA) BlaI family transcriptional regulator, methicillin resistance K02546 regulatory protein (mecI) K02547 methicillin resistance protein (mecR1)

3.6. Statistical Analysis Statistical analysis was performed using the Statistica software (StatSoft, ver.13.3). Error margin ranges represent standard errors of the mean and were calculated by di- viding the standard deviation by the square root of the sample size. The Wilcoxon rank sum test (p < 0.05) was used to conduct multiple nonparametric pairwise compar- isons after Kruskal–Wallis rank sum tests (p < 0.05), to compare the differences among bacterial metapopulations.

4. Conclusions Microorganisms that developed mechanisms which allow them to survive in the presence of an antimicrobial agent, using the possibilities offered by horizontal gene transfer, can colonize other organisms or transfer their resistance genes to other bacteria. The development of new bioinformatics tools allows for a more accurate description of the diversity of the milk microbiota and the association of taxonomic, physiological and functional characteristics, which allows the dairy industry to improve the quality of milk and dairy products. In the presented research, it was found that the species composition of the milk microbiome from three different farms differed significantly. This differentiation was confirmed by the predicted functional analysis of the genetic potential of the bacteria responsible for their metabolism. However, such differences were not shown by the prediction analysis of genes responsible for antibiotic resistance. Therefore, it can be concluded that a reservoir of genes which determine the resistance to antibiotics is supposed to exist and maintained in the population of microorganisms. There is a high probability that bacteria present in milk constitute an excellent reservoir of resistance genes for potentially harmful bacteria, which poses a serious threat to humans and a medical challenge. Furthermore, the increased possibility of the presence of resistant bacteria in the gut flora results in a higher probability of transferring the resistance genes to (potentially) pathogenic bacteria, as well as their distribution in the environment, and their distribution from animals to food of animal origin.

Supplementary Materials: The following are available online, Figure S1: Rarefaction curve used to calculate alpha-diversity, Table S1: The results of read-ﬁltering in sequencing data analysis. Molecules 2021, 26, 5029 11 of 12

Author Contributions: Conceptualization, B.H. and A.P.-C.; methodology, B.H. and J.C. software, J.S.-P. and Ł.W.; validation, B.H. and J.C.; formal analysis, B.H. and A.P.-C.; investigation, B.H., J.S.-P., A.P.-C., P.C., Ł.W. and J.C.; resources, B.H.; data curation, B.H.; writing—original draft preparation, B.H., J.C.; writing—review and editing, B.H. and J.C.; visualization, B.H.; supervision, A.P.-C., J.C.; project administration, A.P.-C., J.C.; funding acquisition, B.H., P.C. and J.C. All authors have read and agreed to the published version of the manuscript. Funding: The publication concerns a project co-financed from European Funds title “Developing Innovative Research Methods For Ensuring The Quality Of Organic, Traditional And Regional Food” under Priority Axis 1 Strengthening innovation and competitiveness of the region’s economy, Measure 1.2 Promoting enterprise investments in research and innovation, Sub-measure 1.2.1 Support for research and development processes, Regional Program Of the Kuyavian-Pomeranian Voivodeship for the years 2014–2020. Publication was co-financed within the framework of the Polish Ministry of Science and Higher Education’s program: “Regional Initiative Excellence” in the years 2019–2022 (No. 005/RID/2018/19)”, financing amount 12,000,000 PLN. The publication was created as a result of the 1st edition of the competition in the program of the Ministry of Science and Higher Education "Implementation doctorate" carried out at the Faculty of Food Sciences and Nutrition, University of Life Sciences in Poznańand at the Institute of Forensic Genetics in Bydgoszcz in 2017–2021. Institutional Review Board Statement: Not applicable. Informed Consent Statement: Not applicable. Data Availability Statement: The 16S rRNA sequencing datasets generated and analyzed during the current study have been deposited at the National Center for Biotechnology Information (SRA reposito-ry), as BioProject under ID PRJNA699887. Conflicts of Interest: The authors declare no conflict of interest. Sample Availability: Not available.

References 1. Skeie, S.B.; Håland, M.; Thorsen, I.M.; Narvhus, J.; Porcellato, D. Bulk tank raw milk microbiota differs within and between farms: A moving goalpost challenging quality control. J. Dairy Sci. 2019, 102, 1959–1971. [CrossRef] 2. Nam, J.H.; Cho, Y.S.; Rackerby, B.; Goddik, L.; Park, S.H. Shifts of microbiota during cheese production: Impact on production and quality. Appl. Microbiol. Biotechnol. 2021, 105, 2307–2318. [CrossRef] 3. Quigley, L.; O’Sullivan, O.; Stanton, C.; Beresford, T.P.; Ross, R.P.; Fitzgerald, G.F.; Cotter, P.D. The complex microbiota of raw milk. FEMS Microbiol. Rev. 2013, 37, 664–698. [CrossRef][PubMed] 4. Addis, M.F.; Tanca, A.; Uzzau, S.; Oikonomou, G.; Bicalho, R.C.; Moroni, P. The bovine milk microbiota: Insights and perspectives from-omics studies. Mol. Biosyst. 2013, 12, 2359–2372. [CrossRef] 5. Wang, Y.; Nan, X.; Zhao, Y.; Jiang, L.; Wang, M.; Wang, H.; Zhang, F.; Xue, F.; Hua, D.; Liu, J.; et al. Rumen microbiome structure and metabolites activity in dairy cows with clinical and subclinical mastitis. J. Anim. Sci. Biotechnol. 2021, 12, 36. [CrossRef] [PubMed] 6. Abebe, R.; Hatiya, H.; Abera, M.; Megersa, B.; Asmare, K. Bovine mastitis: Prevalence, risk factors and isolation of Staphylococcus aureus in dairy herds at Hawassa milk shed, South Ethiopia. BMC Vet. Res. 2016, 12, 270. [CrossRef] 7. Bhakat, C.; Mohammad, A.; Mandal, D.K.; Mandal, A.; Rai, S.; Chatterjee, A.; Ghosh, M.K.; Dutta, T.K. Readily usable strategies to control mastitis for production augmentation in dairy cattle: A review. Vet. World 2020, 13, 2364–2370. [CrossRef][PubMed] 8. De Vliegher, S.; Fox, L.K.; Piepers, S.; McDougall, S.; Barkema, H.W. Invited review: Mastitis in dairy heifers: Nature of the disease, potential impact, prevention, and control. J. Dairy Sci. 2012, 95, 1025–1040. [CrossRef][PubMed] 9. Fusco, V.; Chiefﬁ, D.; Fanelli, F.; Logrieco, A.F.; Cho, G.-S.; Kabisch, J.; Böhnlein, C.; Franz, C.M.A.P. Microbial quality and safety of milk and milk products in the 21st century. Compr. Rev. Food Sci. Food Saf. 2020, 19, 2013–2049. [CrossRef] 10. Rocha, D.C.; Rocha, C.S.; Davi, S.T.; Calado, S.L.M.; Gomes, M.P. Veterinary antibiotics and plant physiology: An overview. Sci. Total Environ. 2021, 767, 144902. [CrossRef] 11. Mohan, H.; Rajput, S.S.; Jadhav, E.B.; Sankhla, M.S.; Sonone, S.S.; Jadhav, S.; Kumar, R. Ecotoxicity, Occurrence, and Removal of Pharmaceuticals and Illicit Drugs from Aquatic Systems. Biointerface Res. Appl. Chem. 2021, 11, 12530–12546. [CrossRef] 12. Ma, Z.; Lee, S.; Fan, P.; Zhai, Y.; Lim, J.; Galvão, K.N.; Nelson, C.; Jeong, K.C. Diverse β-lactam antibiotic-resistant bacteria and microbial community in milk from mastitic cows. Appl. Microbiol. Biotechnol. 2021, 105, 2109–2121. [CrossRef] 13. Church, N.A.; McKillip, J.L. Antibiotic resistance crisis: Challenges and imperatives. Biologia 2021, 76, 1535–1550. [CrossRef] 14. Altamirano, F.L.G.; Barr, J.J. Phage therapy in the postantibiotic era. Clin. Microbiol. Rev. 2019, 32, e00066-18. [CrossRef] 15. van Teeseling, M.C.F.; de Pedro, M.A.; Cava, F. Determinants of Bacterial Morphology: From Fundamentals to Possibilities for Antimicrobial Targeting. Front. Microbiol. 2017, 8, 1264. [CrossRef] Molecules 2021, 26, 5029 12 of 12

16. Gullberg, E.; Cao, S.; Berg, O.G.; Ilbäck, C.; Sandegren, L.; Hughes, D.; Andersson, D.I. Selection of resistant bacteria at very low antibiotic concentrations. PLoS Pathog. 2011, 7, e1002158. [CrossRef] 17. McInnes, R.S.; McCallum, G.E.; Lamberte, L.; van Schaik, W. Horizontal transfer of antibiotic resistance genes in the human gut microbiome. Curr. Opin. Microbiol. 2020, 53, 35–43. [CrossRef] 18. Tilocca, B.; Costanzo, N.; Morittu, V.M.; Spina, A.A.; Soggiu, A.; Britti, D.; Roncada, P.; Piras, C. Milk microbiota: Characterization methods and role in cheese production. J. Proteom. 2020, 210, 103534. [CrossRef][PubMed] 19. Ganda, E.K.; Bisinotto, R.S.; Lima, S.F.; Kronauer, K.; Decter, D.H.; Oikonomou, G.; Bicalho, R.C. Longitudinal metagenomics profiling of bovine milk to assess the impact of intramammary treatment using a third-generation cephalosporin. Sci. Rep. 2016, 6, 37565. [CrossRef] 20. Oikonomou, G.; Machado, V.S.; Santisteban, C.; Schukken, Y.H.; Bicalho, R.C. Microbial diversity of bovine mastitic milk as described by pyrosequencing of metagenomic 16s rDNA. PLoS ONE 2012, 7, e47671. [CrossRef][PubMed] 21. Kuehn, J.S.; Gorden, P.J.; Munro, D.; Rong, R.; Dong, Q.; Plummer, P.J.; Phillips, G.J. Bacterial community profiling of milk samples as a means to understand culture-negative bovine clinical mastitis. PLoS ONE 2013, 8, e61959. [CrossRef] 22. Wouters, J.T.M.; Ayad, E.A.E.; Hugenholtz, J.; Smit, G. Microbes from raw milk for fermented dairy products. Int. Dairy J. 2012, 12, 91–109. [CrossRef] 23. Young, W.; Hine, B.C.; Wallace, O.A.M.; Callaghan, M.; Bibiloni, R. Transfer of intestinal bacterial components to mammary secret ions in the cow. PeerJ 2015, 3, e888. [CrossRef][PubMed] 24. Yuan, L.; Sadiq, F.A.; Liu, T.J.; Li, Y.; Gu, J.S.; Yang, H.Y.; He, G.Q. Spoilage potential of psychrotrophic bacteria isolated from raw milk and the thermo-stability of their enzymes. J. Zhejiang Univ. Sci. B 2018, 19, 630–642. [CrossRef][PubMed] 25. Oultram, J.W.H.; Ganda, E.K.; Boulding, S.C.; Bicalho, R.C.; Oikonomou, G. A Metataxonomic Approach Could Be Considered for Cattle Clinical Mastitis Diagnostics. Front. Vet. Sci. 2017, 4, 36. [CrossRef][PubMed] 26. Masoud, W.; Vogensen, F.K.; Lillevang, S.; Abu Al-Soud, W.; Sørensen, S.J.; Jakobsen, M. The fate of indigenous microbiota, starter cultures, Escherichia coli, Listeria innocua and Staphylococcus aureus in Danish raw milk and cheeses determined by pyrosequencing and quantitative real time (qRT)-PCR. Int. J. Food Microbiol. 2012, 153, 192–202. [CrossRef][PubMed] 27. Bhatt, V.D.; Ahir, V.B.; Koringa, P.G.; Jakhesara, S.J.; Rank, D.N.; Nauriyal, D.S.; Kunjadia, A.P.; Joshi, C.G. Milk microbiome signatures of subclinical mastitis-affected cattle analysed by shotgun sequencing. J. Appl. Microbiol. 2012, 112, 639–650. [CrossRef] [PubMed] 28. Hoque, M.N.; Istiaq, A.; Clement, R.A.; Sultana, M.; Crandall, K.A.; Siddiki, A.Z.; Hossain, M.A. Metagenomic deep sequencing reveals association of microbiome signature with functional biases in bovine mastitis. Sci. Rep. 2019, 9, 13536. [CrossRef] [PubMed] 29. Lima, S.F.; Bicalho, M.L.S.; Bicalho, R.C. Evaluation of milk sample fractions for characterization of milk microbiota from healthy and clinical mastitis cows. PLoS ONE 2018, 13, e0193671. [CrossRef] 30. Pyatov, V.; Vrtková, I.; Knoll, A. Detection of selected antibiotic resistance genes using multiplex PCR assay in mastitis pathogens in the Czech Republic. Acta Vet. Brno 2017, 86, 167–174. [CrossRef] 31. Molineri, A.I.; Camussone, C.; Zbrun, M.V.; Suárez Archilla, G.; Cristiani, M.; Neder, V.; Calvinho, L.; Signorini, M. Antimicrobial resistance of Staphylococcus aureus isolated from bovine mastitis: Systematic review and meta-analysis. Prev. Vet. Med. 2021, 188, 105261. [CrossRef][PubMed] 32. Oliver, S.P.; Murinda, S.E. Antimicrobial resistance of mastitis pathogens. Vet. Clin. N. Am. Food Anim. Pract. 2012, 28, 165–185. [CrossRef] 33. Zeng, X.; Lin, J. Beta-lactamase induction and cell wall metabolism in Gram-negative bacteria. Front. Microbiol. 2013, 4, 128. [CrossRef][PubMed] 34. Quast, C.; Pruesse, E.; Yilmaz, P.; Gerken, J.; Schweer, T.; Yarza, P.; Peplies, J.; Glöckner, F.O. The SILVA ribosomal RNA gene database project: Improved data processing and web–based tools. Nucleic Acids Res. 2013, 41, 590–596. [CrossRef][PubMed] 35. Langille, M.G.; Zaneveld, J.; Caporaso, J.G.; McDonald, D.; Knights, D.; Reyes, J.A.; Clemente, J.C.; Burkepile, D.E.; Vega, T.R.L.; Knight, R.; et al. Predictive functional profiling of microbial communities using 16S rRNA marker gene sequences. Nat. Biotechnol. 2013, 31, 814–821. [CrossRef][PubMed] 36. Dhariwal, A.; Chong, J.; Habib, S.; King, I.; Agellon, L.B.; Xia, J. MicrobiomeAnalyst-a web-based tool for comprehensive statistical, visual and meta-analysis of microbiome data. Nucleic Acids Res. 2017, 45, 180–188. [CrossRef]